Editing Voice Principle (section)

=='''Two Theoretical Perspectives'''==
==='''Cognitive Load Theory'''===
*[https://ectwiki.online/index.php?title=Cognitive_Load_Theory Cognitive Load Theory]: a cognitive architecture of a human being is divided into three parts: a limited working memory, a limitless long-term memory, and schemas that work to organize in long-term memory (Sweller, 2011)

According to cognitive load researchers (Paas & Sweller, 2014), synthetic voices increase extraneous cognitive load and reduce usable cognitive capacity to integrate new information with existing knowledge. An unappealing voice may cause learners more time to generate the information which is more likely to increase the working memory that is distributed into the cognitive thinking process that raises the cognitive load. 

According to cognitive load theory assumptions, the human brain receives instruction in two different channels, verbal and visual, before information processing begins, and the capacity is relatively limited. As a result, synthetic machine-generated voice types may increase the extraneous cognitive load of those exposed to multimedia instruction or engaged in multimedia material because it appeals to the uninterested and distracting in the absence of sufficient social cues (Wouters et al., 2008).

==='''Social Agency'''===
Learning involves social activity(Bandura, 1969). According to social agency theorists (Atkinson et al., 2005), the human voice can be identified quickly due to social interaction and familiarity, resulting in active learning. In summary, social agency theory holds that using social cues in multimedia learning improves educational quality and increases retention (Dinçer & Doğanay, 2017). Social agency theory is a set of ideas that explains how social factors affect multimedia learning (Linek et al., 2010). Cues, including the voice or image of presenters integrated into a multimedia lesson, might act as social stimuli. 

Cues, such as the voice or image of presenters embedded in a multimedia lesson, can serve as social stimuli. The extent to which cues convey social concepts, in particular, can vary. For example, a machine-synthesized voice does not carry the same degree of social cues as the human voice (Mayer et al., 2003). 

While enthusiastic voice has shown that the enthusiastic voice prompted more effective social ratings, the calm voice led to a higher germane load. Furthermore, the embedded social elements give the impression that multimedia instruction involves social interaction rather than a passive lecturing. This may encourage learners to exert the same effort as when interacting with humans. 

In terms of gender differences in voice, Linek et al. (2010) found that the female voice was more effective than the male voice at capturing learners' attention and retention scores. Additionally, the social ratings of the female voice were found to be more assertive and appealing.